Recent advancements in the field of light-field photography have led to the development of high-end plenoptic imaging devices that capture light-field (LF) data. Such imaging devices capture the LF data associated with multiple subjects and/or objects present in a field-of-view (FOV) of the imaging device. Typically, for each object, the LF data corresponds to a vector function that describes the amount of light that flows in every direction through every point in the FOV of the imaging device. The vector function may correspond to a multidimensional (such as five-dimensional or more) plenoptic function that specifies variables, such as a position and a direction of each light ray, with respect to an optical axis. The plenoptic function may exhibit higher dimensionality based on additional variables, such as time, wavelength, and polarization angles. Such higher dimensional plenoptic functions may enable reconstruction of every possible view of the multiple subjects and/or objects, at every moment, from every position, at every wavelength in the FOV of the imaging device.
In certain scenarios, plenoptic imaging devices may generate panoramic images based on multiple FOVs with minimum parallax error. The LF data associated with multiple FOVs may be stitched together to generate panoramic images. However, such a panoramic image may correspond to a flat two dimensional or three dimensional image. Also, the generation of such panoramic images may require substantial memory to store the LF data associated with the multiple FOVs. It may be desired that the generated panoramic images provide a depth perception of the captured subjects and/or objects present in the selected FOVs with optimized memory utilization.
Further limitations and disadvantages of conventional and traditional approaches will become apparent to one of skill in the art, through comparison of described systems with some aspects of the present disclosure, as set forth in the remainder of the present application and with reference to the drawings.